Mechanics and Dynamics of Pinning Points on the Shirase Coast, West Antarctica
Total Page:16
File Type:pdf, Size:1020Kb
https://doi.org/10.5194/tc-2020-298 Preprint. Discussion started: 28 October 2020 c Author(s) 2020. CC BY 4.0 License. Mechanics and dynamics of pinning points on the Shirase Coast, West Antarctica Holly Still1 and Christina Hulbe1 1National School of Surveying, University of Otago, Dunedin, New Zealand Correspondence: Holly Still ([email protected]) Abstract. Ice rises and rumples, sites of localised ice shelf grounding, contribute to shelf-wide mechanics by generating lateral and basal shear stresses, upstream compression and downstream tension. Here, using a case study approach, we simulate Ross Ice Shelf (RIS) and tributary ice stream flow, with and without the Shirase Coast Ice Rumples (SCIR), to quantify the specific contribution of these features to flow dynamics. While longitudinal stretching, and thus ice velocity, increases in response 5 to pinning point removal, flow resistance generated by other grounded features also increases, providing a control on the magnitude of the velocity difference. Spatial variation in two parameters inferred during model initialisation, basal friction and ice softness, further condition the system response to the SCIR. MacAyeal Ice Stream (MacIS), located directly upstream of the SCIR, is less responsive to the loss of the ice rumples than the obliquely oriented Bindschadler Ice Stream due to zones of locally higher basal drag acting on the main trunk of MacIS. In the model, the larger basal drag acting on MacIS is itself, 10 via regional changes in driving stress, a consequence of the coupled ice shelf and ice stream response to the SCIR. We also find that inversion of present-day flow and thickness for basal friction and ice softness, without feature-specific tuning, leads to the incorrect representation of ice rumple morphology, and by extension, any parameter that is affected by the initialisation procedure. Where pinning point effects are important, model tuning that respects pinning point morphology is necessary to represent the ice sheet–ice shelf system as a whole. Viewed from the perspective of change detection, we find that changes to 15 the ice shelf geometry following removal of the SCIR are transient, as mass flux reduces thickness gradients in some areas and increases them in others, while changes to the ice streams persist, even without sustained grounding line retreat. 1 Introduction Ice shelves regulate the Antarctic contribution to sea level rise via their influence on grounding line position and tributary glacier dynamics. Individual ice shelves are regulated by environmental conditions including their geographic setting. An ice 20 shelf laterally confined within an embayment experiences reduced longitudinal tensile stress (and stretching) relative to an unconfined ice shelf due to lateral shearing where the ice flows past coastal features and islands (Sanderson, 1979; Haseloff and Sergienko, 2018). Where floating ice runs aground, a pinning point forms and resulting compression and shearing further reduce longitudinal stresses (Favier et al., 2012; Borstad et al., 2013; Favier and Pattyn, 2015; Berger et al., 2016). Altogether, the rate of mass flux is moderated in an effect commonly referred to as ‘flow buttressing’, a normal force exerted on upstream 25 grounded ice by the ice shelf (Dupont and Alley, 2005, 2006; Gudmundsson, 2013; Fürst et al., 2016). 1 https://doi.org/10.5194/tc-2020-298 Preprint. Discussion started: 28 October 2020 c Author(s) 2020. CC BY 4.0 License. The importance of pinning points to ice shelf stability and grounding line position has been widely examined observationally (e.g., Matsuoka et al., 2015). For example, the speed-up and grounding line retreat of Pine Island Glacier following loss of a pinning point has been documented by Bindschadler (2002), Rignot (2002), Jenkins et al. (2010) and Arndt et al. (2018), and the individual force balance contributions of various Antarctic pinning points have been computed from observational data by 30 Thomas (1973), Thomas (1979), Thomas and MacAyeal (1982), MacAyeal et al. (1987) and Still et al. (2019). More theoretical approaches examine coupled mass and momentum effects of pinning points across the interconnected ice shelf and ice sheet system. For example, Goldberg et al. (2009) conducted idealised simulations of grounding line position and mechanics with (and without) an ice rise to demonstrate how an ice rise can modify vulnerability to the marine ice sheet instability. Favier et al. (2012) demonstrated that local changes to ice thickness due to the emergence of a pinning point generate feedbacks 35 in the stress balance that can maintain the local thickness perturbation and thus the grounded feature. Fried et al. (2014) examined the emergence of ice rises in the Ross Ice Shelf (RIS) as a source of thickness transients that drove past grounding line transgression. Nias et al. (2016) simulated the Thwaites Glacier response to changing contact with a pinning point beneath its floating ice tongue and concluded that basal traction on the grounded ice was more important to the glacier response than the direct mechanical effects of the pinning point itself. All of these examples demonstrate the importance of non-local effects 40 of pinning points in the ice shelf and ice sheet system. By generating resistive stresses, pinning points modify the velocity pattern and, via advection, the thickness pattern of an ice shelf. The momentum and mass perturbations together must be balanced by changes in thickness and resistive stresses elsewhere in the ice shelf–ice sheet system. Enhanced deformation around a pinning point also affects ice properties such as ice crystal fabric and temperature, modifying softness and thus ice flow (e.g., Borstad et al., 2013). The aim of the present work 45 is to examine the complete system of mass and momentum adjustments using a case study approach for a collection of pinning points in the RIS, West Antarctica. The flow of the RIS is regulated by a diverse collection of ice rises and rumples, the surface morphological expressions of ice shelf flow over or around pinning points (Fig. 1). Ice rises have a distinct dome-shaped morphology with ice shelf flow diverging around the rise, while undulating ice rumples form where the ice continues to flow directly over the area of localised 50 grounding (Martin and Sanderson, 1980; Matsuoka et al., 2015). Large ice rises in the RIS include Crary Ice Rise, Roosevelt Island and Steershead Ice Rise. Smaller, unnamed ice rumples are located along the Siple and Shirase Coasts. A group of these ice rumples in the eastern RIS, which we refer to collectively as the Shirase Coast Ice Rumples (SCIR) were chosen for this study. While individual rumples in the SCIR complex are relatively small, the lightly–grounded and low basal traction features collectively generate flow resistance comparable to larger ice rises in the RIS (Crary Ice Rise, Steershead Ice Rise 55 and Roosevelt Island) (Still et al., 2019). The SCIR are located 60 km downstream from the grounding lines of the MacAyeal and Bindschadler Ice Streams (MacIS and BIS), two large outlet streams of the West Antarctic Ice Sheet, and within a cove upstream of Roosevelt Island. This setting stands out as a region of significant buttressing in the RIS (Reese et al., 2018) and the arrangement allows the mass and momentum balances to be examined in a non-simple geometry. Observational-data driven analysis provides a limited, snapshot-like view of recent conditions. Observed velocity and thick- 60 ness may be used directly to quantify effective flow resistance, for example, using a force budget approach (MacAyeal et al., 2 https://doi.org/10.5194/tc-2020-298 Preprint. Discussion started: 28 October 2020 c Author(s) 2020. CC BY 4.0 License. Figure 1. Pinning points in the RIS and the model domain boundary. Large pinning points are labelled: SCIR = the Shirase Coast Ice Rumples, RI = Roosevelt Island, SIR = Steershead Ice Rise and CIR = Crary Ice Rise. The colour map of surface ice velocity magnitude is from the MEaSURES velocity dataset (Rignot et al., 2011a). The black line indicates the grounding zone (Bindschadler et al., 2011). In each figure from hereon, datasets are mapped with a Polar Stereographic Projection with a central meridian of 0◦ and a standard latitude of 71◦S, and in most cases, overlayed onto the MODIS MOA (Haran et al., 2014). 1987; Still et al., 2019) but such calculations cannot answer questions about the non-local response. Flow buttressing numbers (Gudmundsson, 2013; Fürst et al., 2016) provide a summary view of the non-local effects but do not quantify the pinning point contribution to individual resistive stresses. We improve upon past work by conducting a modelling experiment in which the RIS and its tributary ice streams are simulated with and without the SCIR complex, using the Ice-sheet and Sea-level System 65 Model (ISSM) (Larour et al., 2012). Differences between the two models quantify the complete dynamical influence of the SCIR on the ice sheet–ice shelf system. Properties of the ice and subglacial bed, inferred during model initialisation, can also be examined using this approach. 2 Method 2.1 Ice sheet model 70 ISSM is an open-source, finite-element ice flow model that solves the conservation equations for mass and momentum in combination with appropriate boundary conditions and the constitutive relationship for ice (Larour et al., 2012). The Shallow Shelf (or shelfy-stream) Approximation (SSA) (Morland, 1987; MacAyeal, 1989) of the full-Stokes equations, appropriate for sliding over a very weak substrate such as water or water-saturated subglacial till, is used to simulate ice shelf and ice sheet flow. The fast flowing RIS ice streams draining the West Antarctic Ice Sheet are characterised by thawed bases and significant 75 sliding over soft subglacial till (MacAyeal et al., 1995; Joughin et al., 2004), justifying the SSA for the present study.